summaryrefslogtreecommitdiff
path: root/text_recognizer/model/greedy_decoder.py
blob: 2c4c16ece5bced22ce195ee555b7336d6e9b9f14 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
"""Greedy decoder."""
from typing import Type
from text_recognizer.data.tokenizer import Tokenizer
import torch
from torch import nn, Tensor


class GreedyDecoder:
    def __init__(
        self,
        network: Type[nn.Module],
        tokenizer: Tokenizer,
        max_output_len: int = 682,
    ) -> None:
        self.network = network
        self.start_index = tokenizer.start_index
        self.end_index = tokenizer.end_index
        self.pad_index = tokenizer.pad_index
        self.max_output_len = max_output_len

    def __call__(self, x: Tensor) -> Tensor:
        bsz = x.shape[0]

        # Encode image(s) to latent vectors.
        img_features = self.network.encode(x)

        # Create a placeholder matrix for storing outputs from the network
        indecies = (
            torch.ones((bsz, self.max_output_len), dtype=torch.long, device=x.device)
            * self.pad_index
        )
        indecies[:, 0] = self.start_index

        for i in range(1, self.max_output_len):
            tokens = indecies[:, :i]  # (B, Sy)
            logits = self.network.decode(tokens, img_features)  # [ B, N, C ]
            indecies_ = torch.argmax(logits, dim=2)  # [ B, N ]
            indecies[:, i] = indecies_[:, -1]

            # Early stopping of prediction loop if token is end or padding token.
            if (
                (indecies[:, i] == self.end_index) | (indecies[:, i] == self.pad_index)
            ).all():
                break

        # Set all tokens after end token to pad token.
        for i in range(1, self.max_output_len):
            idx = (indecies[:, i - 1] == self.end_index) | (
                indecies[:, i - 1] == self.pad_index
            )
            indecies[idx, i] = self.pad_index

        return indecies